Development of an Automatic Landmine Detection and Marking System for the Demining Robot Gryphon

Abstract

Mechanical systems or robots are designed to support human operators during complex and dangerous tasks such as demining operations. Even though the robot Gryphon was created to automate these operations, some of its tasks still rely greatly on the human operator, who has few or no assisting tools to perform efficient decisions. During the landmine detection and marking task in special, the operator is totally responsible for analyzing the scanned data and pointing the potential targets, which makes the system performance unstable and vulnerable to human factors. This article proposes an automatic method for finding potential targets, which the operator has the simple role of accepting or not the decisions taken by the automatic method. Experimental results showed that time duration, POD and FAR were greatly improved compared to the former methods.